A method for electric heating flexible resource light robust collaborative optimization configuration

Abstract

The application provides a light-robust collaborative optimization configuration method of electric heating flexible resource, analyzes the flexible resource improvement mechanism of various resources from the aspects of meeting system peak regulation and standby requirements, and refines the collaborative planning mechanism from the aspects of regulation performance and economy. Considering the influence of wind power uncertainty on the economy and safety of the investment scheme, a three-layer light-robust collaborative planning model is established, which takes the minimum sum of three costs, i.e. incremental investment in the planning target year, operation and standby deficiency risk, as the target, contains investment decision and day-ahead deterministic optimization scheduling of the main problem and day-to-day equipment rescheduling and standby deficiency risk evaluation of the sub-problem considering wind power uncertainty, and adopts strong duality theory and CCG algorithm for main and sub-problem iterative solution.

Description

Technical Field

[0001] This invention belongs to the field of electrical engineering technology, and mainly relates to a method for lightweight and robust collaborative optimization of electrothermal flexibility resources. Background Technology

[0002] Renewable energy sources such as wind and solar power will see further development, significantly increasing their share in the power structure and the need for system flexibility. Thermal power units, due to their high stability, flexibility, and large scale, have always been the most important flexible regulation resource in the power system. The "Three Norths" region (Northeast, North, and Northwest my country) is a large-scale wind power base in China. Due to its high latitude and long winter heating season, it has a high annual heat load demand, with thermal power units accounting for as much as 80% of thermal power. During the heating season, the passive increase in thermal power output due to thermoelectric coupling leads to insufficient peak-shaving capacity of the system when wind power generation is high, resulting in wind curtailment. In 2022, the wind curtailment rate in eastern Inner Mongolia reached 10%. Current solutions include flexible retrofitting of thermal power units, such as high back-pressure retrofitting and low-pressure cylinder removal retrofitting, or configuring thermal storage, electric boilers, and demand-side response to increase the system's peak-shaving capacity while ensuring heat load supply. However, due to differences in investment economics, peak-shaving improvement, operational economics, scenario adaptability, and the ability to provide system backup among various measures and resources, how to achieve synergistic efficiency through matching and optimization is an urgent problem to be solved.

[0003] Research on flexibility retrofitting and resource allocation for systems with a high proportion of renewable energy has yielded substantial results. Research scenarios can be broadly categorized into two types: one involves coordinated planning of flexibility resources and power sources. Some scholars, considering the differences in user participation time and response volume, have proposed a two-way guided tiered incentive mechanism for demand-side response resources and established a supply-demand master-slave game-theoretic power planning model. This method significantly reduces the total system cost and promotes renewable energy consumption. Other scholars have proposed a method for coordinating flexibility retrofitting of thermal power units with power sources, establishing a robust optimization planning model with the goal of minimizing investment and operating costs. Numerical examples verify that the overall economic efficiency of coordinated planning is significantly better than separate power source planning and coordinated planning of energy storage and power sources. The other type of research focuses on the optimal allocation of flexibility resources for existing systems. Some studies, considering the differentiated characteristics of flexibility retrofitting for pure condensing units and combined heat and power units, have proposed a two-stage retrofitting planning model that considers the game-theoretic investment decisions of multiple power generation companies. This method can balance the interests of power generation companies and the feasibility of the plan, but reduces the scale of the retrofit. A study has established a two-layer optimization model that coordinates various measures. The upper layer optimizes the capacity of flexible resources such as deep peak-shaving retrofits of cogeneration units, peak shaving of self-owned power plants, demand response, thermal storage, and energy storage. The lower layer simulates optimized production of the system's electricity and heat during the target year. The study points out that as the proportion of renewable energy increases, the capacity for cogeneration peak shaving retrofits increases linearly. Energy storage, due to its high cost, is only technically and economically viable with an extremely high proportion of renewable energy, and its demand response regulation capacity is limited, resulting in a negligible increase. Therefore, compared to indirect decoupling measures such as energy storage and demand response, fully leveraging the flexibility of cogeneration peak shaving is the most direct, economical, and effective method to address insufficient peak shaving caused by cogeneration coupling.

[0004] Some studies have addressed the configuration of thermal power units with thermal storage and electric boilers, proposing a multi-objective programming method aimed at maximizing total capacity and minimizing system operating costs. The research indicates that thermal storage is more economical than electric boilers in terms of investment, and is more suitable for low-wind and high-wind scenarios respectively. Coordinating the configuration of both improves investment efficiency. However, simulations only consider typical days during the heating season and do not account for the impact of the configuration scheme on the annual operating economy, resulting in low investment utilization. From the perspective of the essence of thermoelectric decoupling, thermal storage reduces heat demand during peak-shaving periods by shifting loads, thereby reducing both heat load and unit power output; electric boilers increase power load and compensate for heating, improving down-peak shaving capacity. Neither changes the thermoelectric coupling characteristics of the unit, but rather assists in achieving peak-shaving through changes in heat and power loads. Furthermore, since thermal storage for heating can be converted into unit power generation reserves, it can be used to reduce the risk of insufficient reserves caused by uncertain source loads. Both of the above are methods of compensatory heating. For unit-side modifications, high back pressure modification and low-pressure cylinder removal increase thermal output and reduce electrical output by raising the back pressure during the heating season, thus achieving partial decoupling and altering the thermoelectric coupling characteristics to some extent. However, they differ in peak-shaving range and economic efficiency. Overall, unit-side modifications offer better investment economics, but have a narrower adjustment range and lower operational flexibility. Coordinating with compensatory heating equipment can achieve synergistic effects. The source-load uncertainty of high-proportion wind power systems has a significant impact on the economic efficiency and operational safety of flexibility modification schemes, but existing research has been limited.

[0005] In China's "Three Norths" region (Northwest, North, and Northeast China), wind resources are abundant, but system flexibility resources are scarce. During the heating season, a high proportion of wind power and a high proportion of combined heat and power (CHP) coexist within the system, crowding out wind power grid connection space and posing a severe challenge to the safe and economical operation of the system. To improve the economic efficiency of wind power consumption, this invention proposes a robust collaborative optimization configuration method for CHP flexibility resources that takes into account the uncertainties of wind power. This method addresses the differences in investment economics, peak-shaving improvement, operational economics, scenario adaptability, and system backup capacity by coordinating and optimizing various measures and resources, such as high back-pressure retrofitting of CHP units, low-pressure cylinder removal retrofitting, and the configuration of thermal storage and electric boilers. Summary of the Invention

[0006] To address the above problems, this invention proposes a lightweight and robust collaborative optimization configuration method for electrothermal flexibility resources, comprising the following steps:

[0007] Step 1: Establish a three-layer, two-stage, lightweight robust programming model that includes investment decision-making, day-ahead deterministic optimization scheduling, and intraday rescheduling that takes into account wind power uncertainty.

[0008] Step 2: Represent the three-layer, two-stage lightweight robust programming model established in Step 1 as a two-stage, three-layer optimization problem of min-max-min;

[0009] Step 3: Use the CCG algorithm to divide the two-stage, three-layer optimization problem in Step 2 into a main problem and sub-problems, and iterate to find the optimal solution. The main problem solves the first-stage variables, namely planning and day-ahead deterministic scheduling schemes. The sub-problems find the worst-case scenario that maximizes the risk of insufficient reserves under given planning and day-ahead scheduling schemes.

[0010] Preferably, the three-layer, two-stage lightweight robust planning model established in step 1 includes, with the annual investment cost C of newly added resources at the planning level as the annual level. INV Annual incremental operating cost C INC-OP Annual reserve shortage risk cost C RISK An investment decision-making model is established with the objective of minimizing the sum of the three parts; based on the typical daily operating cost of n. A day-ahead deterministic optimization scheduling model is established for the objective; an intraday rescheduling model is established considering the uncertainty of wind power, whose objective function includes risk penalties represented by wind curtailment and load shedding penalties, and the cost of new unit generation; the annual investment cost C of the new resources. INV It consists of four types of flexible transformation methods and resource costs; the annual incremental operating cost C INC-OP It is the difference in annual operating costs before and after the renovation; the annual reserve shortage risk cost C RISK The calculation formula is:

[0011]

[0012] In the formula, β represents the risk penalty preference. The typical daily reserve shortage risk cost consists of two parts: intraday adjustment costs and risk penalties; N type Typical number of days; d n This represents the number of typical operating days (n).

[0013] Specifically, the four types of flexible retrofitting methods include low-pressure cylinder removal retrofitting, high back pressure heating retrofitting, configuration of thermal storage, and configuration of electric boilers; the difference between the compensation heating cost of each resource and the original unit's direct co-production heating cost is used as the measurement standard for peak-shaving cost, with coal consumption as its unit of measurement.

[0014] The low-pressure cylinder cut-off modification refers to cutting off the steam inlet of the low-pressure cylinder of the extraction steam unit, keeping only a small amount of cooling steam, so that the low-pressure cylinder operates under high vacuum conditions with "zero output". The power generation reduction and the heating power increase after the cut-off are constant values, with double the peak-shaving space. When the cylinder is cut off, the steam used to do work in the low-pressure cylinder is completely extracted for heating, which is still cogeneration heating, and the peak-shaving cost is 0.

[0015] The high back pressure heating retrofit refers to the turbine operating at a higher back pressure during the heating season, using the exhaust steam discharged after generating electricity from the high, medium, and low pressure cylinders to heat the heating network water. The heating network water then enters the heating network heater where it is supplemented with extracted steam for heating, thus achieving decoupling of the heat and electricity components. This enhances the heating peak-shaving capacity and reduces the power supply regulation range. The exhaust steam increases the unit's heating capacity. According to the "benefit goes to electricity" method, the high back pressure heating peak-shaving cost is negative.

[0016] The described thermal storage configuration involves connecting the thermal storage device in parallel with the heating unit. When there is surplus heat from the combined heat and power generation of the steam turbine, the thermal storage is stored. When the power output of the steam turbine needs to be reduced or increased for peak shaving, the thermal storage is released to supplement the insufficient heat supply from the steam turbine. This configuration also enhances peak shaving capabilities and provides backup capacity to cope with the uncertainty of wind power in extreme situations. The heat supply for compensation comes from the combined heat and power generation during other periods, and the peak shaving cost is 0.

[0017] The electric boiler configuration involves connecting it in parallel with the original heating system of the unit. When the electric boiler operates at full load, the overall power consumption decreases while the heating power increases. When the system's peak shaving is insufficient, the electric boiler is started to consume electricity for heating. The unit's heating demand decreases, causing its power output to decrease, thus achieving double peak shaving and maximizing peak shaving costs.

[0018] Preferably, the constraints after the low-pressure cylinder removal modification include:

[0019] 1) Feasible operating range constraints:

[0020]

[0021]

[0022]

[0023] In the formula: 0-1 variables and γ LC,i,t These represent the start-up and cylinder-off status of the extraction steam turbine unit during time period t within day i, respectively. and These represent the combined heat and power (CHP) power supply and heat power components of the low-pressure cylinder cut-off retrofitted unit in the non-cut-off and cut-off cylinder sections, respectively. This indicates the overall electrothermal output of the unit;

[0024] 2) Daily low-pressure cylinder cutoff frequency constraint

[0025]

[0026] In the formula: T is the intraday scheduling cycle;

[0027] The constraints after the high back pressure heating system renovation are:

[0028] 1) Modify the start-up operating range constraints:

[0029]

[0030] In the formula: 0-1 variables and These refer to whether the extraction steam turbine unit has undergone high back pressure heating retrofit and its operating status, respectively. m1,i and c m2,i These are the electrothermal ratios of extraction back-pressure turbine unit i when it operates under pure back-pressure and maximum extraction conditions, respectively. and These are the maximum and minimum generating capacities of unit i under pure back pressure conditions, respectively. This represents the maximum heating power of unit i.

[0031] 2) Unit ramp rate constraint: The ramp rate constraint of back-pressure units is represented by the power ramp rate constraint under pure back-pressure conditions with a slope of -1 on the equal coal consumption line.

[0032]

[0033] In the formula: and These are the uphill and downhill ramp rates for unit i, respectively. The electrical and thermal power of unit i during time period t is converted to the electrical power under pure back pressure condition under the equal coal consumption line;

[0034] 3) Unit coal consumption function: The coal consumption of a back-pressure unit is expressed by the coal consumption under pure back-pressure operating conditions on an equal coal consumption line.

[0035]

[0036] The thermal storage charging and discharging power and energy constraints of the configured thermal storage method are as follows:

[0037]

[0038]

[0039] In the formula: and The heat storage capacity and maximum charging and releasing capacity of the thermal storage unit i during the specified time period t; η HA For heat storage efficiency; For capacity, and These are the maximum heat storage and release power of the thermal storage device;

[0040] The operating constraints of the electric boiler are:

[0041]

[0042] In the formula: C represents the power consumption of the boiler supplied to unit i during time period t. EB η EB This refers to the capacity and electro-thermal conversion efficiency of the electric boiler.

[0043] Preferably, the typical daily n-day operating cost in the day-ahead deterministic optimization scheduling model is... The model optimizes the power generation costs of extraction steam and pure condensing units, including coal consumption and carbon emissions, using the following variables: output and operating status of pure condensing unit i during a typical day, time period n. and The startup status indicators of the extraction steam turbine unit, its heating and power generation output, and the cylinder cut-off status of the unit after the low-pressure cylinder removal modification. and γ LC,i,n,t The extraction back-pressure unit's heating and power generation output and start-up status. and And the thermal storage output, thermal storage capacity and electric boiler output of extraction steam unit i and and wind power output The objective function is then expressed as:

[0044]

[0045]

[0046]

[0047] In the formula: ρ is the unit coal price; ρ is the equivalent coal price calculated from water, environmental protection, and other costs excluding coal consumption during unit operation, which is 1.15; τ is the length of a unit time period, which is 1 hour; γ car Ω represents the unit carbon emission cost; θ is the standard coal CO2 emission coefficient, taken as 0.68; PU For pure condensing units; T is the typical number of hours in a day, which is 24 hours; F Υ,i,n,t and These represent the coal consumption for power generation and the coal consumption for start-up, respectively. Υ={CHP-Λ,CON} is the set of unit types, and Λ={φ,B} is the set of various extraction steam units.

[0048] Preferably, the day-ahead deterministic optimization scheduling model includes operating constraints for pure condensing units, operating constraints for extraction steam units, operating constraints after the configuration of four types of flexibility modification measures, heating balance constraints for each heating zone, system power supply balance constraints, and wind power output constraints.

[0049] Preferably, the operating constraints of the pure condensing unit include limit output, ramp-up, shortest start-up and shutdown time constraints, and coal consumption function.

[0050]

[0051]

[0052]

[0053]

[0054] In the formula, and These represent the minimum and maximum output of unit i, the upper and lower limit ramp rates, and the shortest start-stop time, respectively.

[0055] The extraction steam turbine units should be fully operational during the heating season, and may participate in start-up and shutdown scheduling during the non-heating season, provided that start-up and shutdown time constraints are met. Additionally, the operational constraints for the extraction steam turbine units include limit ramp-up constraints and a coal consumption function.

[0056]

[0057]

[0058]

[0059]

[0060] In the formula, M sea This is a seasonal indicator; it is 1 for the heating season and 0 for the non-heating season. and These represent the uphill and downhill ramp rates of extraction steam turbine unit i under pure condensing conditions; a i and b i These are the marginal coal consumption rate and no-load coal consumption of extraction steam turbine unit i under pure condensing conditions, respectively.

[0061] Preferably, the heat supply balance constraint of each heating zone is:

[0062]

[0063] In the formula, Ω EX,l This refers to the collection of all heating units in a thermal power plant. The heat load borne by a thermal power plant during a typical day n-hour period t.

[0064] Preferably, the system power supply balance constraint is:

[0065]

[0066] In the formula: and This represents the wind power output and system electrical load during a typical day, n-time period t.

[0067] Preferably, the objective function of the intraday rescheduling model, which takes into account wind power uncertainties, is:

[0068]

[0069]

[0070] C wp C loss , and The decision variable x includes the unit power wind curtailment and load shedding penalty cost, as well as the amount of wind curtailment and load shedding. In addition to wind curtailment and load shedding, the decision variable x also includes intraday rescheduling variables for various types of generating units, thermal storage, and electric boilers.

[0071] Preferably, the intraday rescheduling model includes four types of operational constraints after the configuration of flexibility modification measures: thermal power balance constraints, electrical power balance constraints, wind curtailment constraints, and emergency load shedding constraints.

[0072] Preferably, the thermal power balance constraint is

[0073]

[0074] The power balance constraint is

[0075]

[0076] The wind curtailment constraint is

[0077]

[0078] The emergency load shedding constraint is as follows:

[0079]

[0080] Preferably, the two-stage, three-layer optimization problem is expressed as follows:

[0081]

[0082]

[0083] In the formula: h(x) is the constraint condition of the first-stage investment-day deterministic scheduling problem, which does not contain uncertain parameters; g(x,u,y) and l(x,u,y) are the inequality and equality constraints of the rescheduling problem under the second-stage uncertain scenario, respectively; Ω(x,u) represents the feasible region of the running variable y given the planning variable x and the uncertainty set u.

[0084] Preferably, in step 3, the two-stage, three-layer optimization problem is decomposed to obtain the objective function and constraints of the main problem:

[0085]

[0086]

[0087] In the formula: c1 and c2 are the coefficients of the decision variables of the objective function in the first and second stages, respectively; k is the current iteration number of the algorithm, and y l The solution to the subproblem after l iterations is the main problem and the subproblem's solution. Let be the value of the uncertain variable u in the worst-case scenario obtained after l iterations;

[0088] The objective function of the decomposed subproblems is:

[0089]

[0090] Based on strong duality theory, the objective function of the subproblem is transformed into a max form, which is:

[0091]

[0092]

[0093] The uncertain wind power output during a typical day n-hour period t is u=u+B + Δu + -B - Δu - Substituting the transformed subproblem objective function into the linearized function using auxiliary variables, we obtain...

[0094]

[0095]

[0096] In the formula: B' + B' - C' + C' - M is a sufficiently large positive real number and is introduced as a continuous auxiliary variable.

[0097] The aforementioned method for lightweight and robust collaborative optimization of electrothermal flexibility resources is characterized in that: the specific process of iterative iteration in step 3 is as follows:

[0098] Step 1) Generate a set of predicted wind power output values ​​u k As the worst-case scenario variable for the model, we set the lower bound of the total cost corresponding to the final planning scheme as LB = –∞, the upper bound as UB = +∞, and the number of iterations as k = 1;

[0099] Step 2) Determine the worst-case scenario Substituting these values ​​into the objective function of the main problem, we obtain the corresponding planning and execution results. k is the iteration number, which is 1 in this case. The calculated comprehensive cost is used as a lower bound of the model and assigned to it.

[0100] Step 3) Substituting the fixed values ​​into the linear expression of the objective function of the decomposed subproblem, we obtain the objective function value under the uncertain scenario of the subproblem. and the value of the uncertain variable u under the corresponding worst-case scenario. Update Upper Realm

[0101] Step 4) Determine if the difference between UB and LB is less than the given convergence threshold ε. If it is less, output the optimal solution at this time. Otherwise, let the iteration number k = k + 1, jump to step 2) to continue solving, and increase the variable y. k+1 With the following constraints:

[0102]

[0103] The beneficial effects of this invention are as follows: 1) The planning model of this invention can effectively ensure the safe supply of thermal power load while promoting wind power consumption and improving overall economic efficiency. Regardless of whether it is the heating season or the non-heating season, by compensating for the thermal storage of heating equipment and coordinating the operation of electric boilers with the modified extraction steam turbines, as well as through day-ahead and intraday two-stage optimization adjustments, the wind power consumption space can be fully and effectively released, improving system economic efficiency. 2) The lightweight robust optimization algorithm proposed in this invention can effectively improve the risk resistance of the planning scheme and reduce conservatism. With optimal economic efficiency as the goal, when wind power uncertainty is low and the risk penalty preference is low, investment is lower, and there is a greater tendency to carry out high back pressure modification and configure electric boilers; when wind power uncertainty is high and the risk penalty preference is high, investment is higher, and there is a greater tendency to carry out low-pressure cylinder removal modification and configure thermal storage. Attached Figure Description

[0104] Figure 1 This is a schematic diagram of the load and electrical / thermal coupling relationship of a heat / electric coupling system with a high proportion of wind power during the heating season in an embodiment of the present invention;

[0105] Figure 2 This is a schematic diagram of the feasible operating range of the thermal / electric coupling system with a high proportion of wind power in this embodiment of the invention after the low-pressure cylinder is removed and the high back pressure is modified;

[0106] Figure 3 This is a schematic diagram of the feasible operating range of the thermal power unit after configuring thermal storage and electric boilers in the thermal / electric coupling system with a high proportion of wind power in this embodiment of the invention;

[0107] Figure 4This is a flowchart of a two-stage lightweight robust planning algorithm for the electrothermal flexibility resources of a thermal / electric coupling system with a high proportion of wind power, as described in an embodiment of the present invention. Detailed Implementation

[0108] The embodiments will now be described in detail with reference to the accompanying drawings.

[0109] This invention first analyzes the mechanism for improving the flexibility of various resources from the perspective of meeting the needs of system peak shaving and reserve, and extracts the collaborative planning mechanism from the perspective of regulation performance and economy. Then, considering the impact of wind power uncertainty on the economy and safety of investment schemes, a three-layer lightweight robust collaborative planning model is established. Its objective is to minimize the sum of three costs: annual incremental investment, operation and reserve shortage risk. It includes the main problem of investment decision and day-ahead deterministic optimization scheduling, and the sub-problems of intraday equipment rescheduling and reserve shortage risk assessment that take into account wind power uncertainty. The main sub-problems are solved iteratively using strong duality theory and CCG algorithm.

[0110] 1. Mechanism of Electrothermal Flexibility Resource Modeling and Collaborative Planning

[0111] 1.1 Flexibility Requirements Analysis of Thermal / Electrical Coupling Systems

[0112] Load and electrical / thermal coupling relationship of a heat / electric coupling system with a high proportion of wind power during the heating season, as follows: Figure 1 As shown. In the left figure, the thick dashed and thick solid curves represent the electrical load of the wind power unit before and after grid connection, respectively. In the right figure, region ABCD represents the operating range of thermal power unit i, c v,i c m,i The slopes of segments AB(DC) and CB are respectively. and These represent the maximum and minimum power generation of the unit under pure condensing conditions, respectively. This represents the unit's maximum heating capacity. Existing cogeneration units operate on a heat-driven power generation principle, meaning the heating output is first determined based on the heat load, and then the unit's adjustable power generation capacity is determined based on the heating output. This applies when the unit's heat load is at a high level during different time periods. At that time, due to the reduction of the electrical load of the heating units by wind power grid connection, the electrical load was lower than that during certain periods (bg period). In order to meet the heating demand, the units have to maintain a high forced output. Wind curtailment, specifically the area curtailed, is shown in the graph (bcdefg). Therefore, improving the flexibility of the thermal / electric coupling system requires increasing thermal output to ensure heat load supply while reducing electrical output, thus freeing up grid connection space for wind power.

[0113] 1.2 Mechanism and Modeling for Enhancing the Flexibility of Multiple Resources

[0114] 1.2.1 Low-pressure cylinder removal and modification

[0115] The low-pressure cylinder cutoff modification method refers to cutting off the steam inlet to the low-pressure cylinder of the extraction steam unit, maintaining only a small amount of cooling steam, so that the low-pressure cylinder operates under high vacuum conditions with "zero output". The operating range is as follows: Figure 2 As shown in (a). After this modification, the low-pressure cylinder can be flexibly disconnected during the day. When not disconnected, it operates in the original ABCDA extraction steam range; when disconnected, it operates in the B'C' section back pressure state range. and These represent the reduced power generation and increased heating power after cylinder cutoff, respectively, and are essentially constant values. Cylinder cutoff operation increases the unit's thermal output and reduces its electrical output (E"E'), causing a decrease in the system's heating demand before the modification, which in turn lowers its electrical output (E"E"'), creating double the peak-shaving capacity (E"E'). Considering that a 1 unit reduction in electricity on the power side may lead to additional coal consumption for heating on the heating side, this invention uniformly uses the difference between the compensation heating cost of each resource and the original unit's direct co-production heating cost as the measurement standard for peak-shaving cost, with coal consumption as its unit of measurement. During cylinder cutoff operation, it is equivalent to operating under back pressure conditions with only high and medium pressure cylinders. The steam originally used for power in the low-pressure cylinder is completely extracted for heating, still belonging to co-production heating, with a peak-shaving cost of 0. The various constraints of the unit change after this modification as follows:

[0116] 1) Feasible operating range constraint: Set 0-1 variables and γ LC,i,t These represent the start-up and cylinder cut-off states of the extraction steam turbine unit during the intraday period t, respectively. and Let A, B, C, D, and E represent the combined heat and power (CHP) components of the low-pressure cylinder cut-off retrofitted unit in the non-cut-off interval ABCDA and the cut-off interval B'C', respectively. Then, the combined heat and power output of the two intervals and the unit as a whole is as follows: The feasible operating intervals are represented as follows:

[0117]

[0118]

[0119]

[0120] 2) Daily Low-Pressure Cylinder Cut-off Frequency Constraint: Due to the risks associated with frequent switching, this invention assumes only one cut-off per day. Since each switch results in two changes to the cylinder cutting status, the constraint is expressed as follows:

[0121]

[0122] In the formula: T is the intraday scheduling cycle.

[0123] 1.2.2 High back pressure heating system renovation

[0124] High back pressure heating retrofit refers to the process where a steam turbine operates at a higher back pressure during the heating season, using the exhaust steam from the high, medium, and low-pressure cylinders to generate electricity and then heating the heating network water. The heating network water then enters the heating network heater where it is supplemented with extracted steam for heating, thus decoupling the thermoelectric components. After the extraction steam unit is retrofitted from a high back pressure unit to an extraction back pressure unit, its feasible operating range is as follows: Figure 2 As shown in MNKLM in (b), the slope is c m1 NK, c m2 The area bounded by the ML line and two iso-coal consumption lines NM and KL with a slope of -1 is located on the lower right side of the pre-modification operating range. Although the heating peak-shaving capacity is enhanced, the power supply regulation range is reduced and narrowed. Since the exhaust steam increases the unit's heating capacity, according to the "benefit to electricity" method, the cost of back-pressure heating is lower than that of extraction steam heating; therefore, the heating peak-shaving cost is negative. The modification also affects the unit's ramp rate and coal consumption. Therefore, the constraints after the high back-pressure modification of the extraction steam unit change as follows:

[0125] 1) Modify the start-up operating range constraints: Set 0-1 variables and If the extraction steam turbine unit i has undergone high back pressure modification and is in operation, then the operating range MNKLM is represented as:

[0126]

[0127] In the formula: c m1,i and c m2,i These are the electrothermal ratios of extraction back-pressure turbine unit i when it operates under pure back-pressure and maximum extraction conditions, respectively. and These are the maximum and minimum generating capacities of unit i under pure back pressure conditions, respectively. This represents the maximum heating power of unit i.

[0128] 2) Unit ramp rate constraint: The ramp rate constraint of back pressure unit can be represented by the power ramp rate constraint under pure back pressure condition under the iso-coal consumption line (slope of -1).

[0129]

[0130] In the formula: and These represent the uphill and downhill ramp rates of unit i, respectively. In equation (6) The electrical and thermal power of unit i during time period t is converted to the electrical power under pure back pressure conditions under the equal coal consumption line.

[0131] 3) Unit coal consumption function: The coal consumption of a back pressure unit can be represented by the coal consumption under pure back pressure conditions under the equal coal consumption line.

[0132]

[0133] 1.2.3 Configure thermal storage

[0134] Configure thermal storage method to have a capacity of The maximum heat storage and release power are respectively and The thermal storage device is connected in parallel with the heating unit i. The overall feasible operating range of the extraction steam unit after configuring thermal storage is shown in [reference needed]. Figure 3 As shown in (a), AA'B'C”C'D', through the charging and releasing of heat in the heat storage tank, causes the entire AB and BC segments to shift to the right by C. HA,f The CD segment is shifted to the left by C. HA,c This method stores residual heat from combined heat and power generation by the steam turbine. When peak power demand requires a decrease (increase) in the steam turbine's output, the heat is released (EE' and FF' segments) to supplement the insufficient heat supply from the steam turbine. This also enhances both peak and off-peak capacity (E”E' and F’F” segments) and provides backup capacity to cope with the uncertainty of wind power in extreme situations. The compensating heat supply comes from the combined heat and power generation during other periods, thus the peak-shaving cost is zero. Storage charging and discharging power and energy constraints:

[0135]

[0136]

[0137] In the formula: and The heat storage capacity and maximum charging and releasing capacity of the thermal storage unit i during the specified time period t; η HA For thermal storage efficiency.

[0138] 1.2.4 Configure an electric boiler

[0139] This method will have a capacity of C EB The electrothermal conversion efficiency is η EB The electric boiler is connected in parallel with the original heating system of the unit. After configuring the electric boiler, the overall feasible operating range of the extraction steam unit is as follows: Figure 3 As shown in (b), for any point within the operating range ABCD, the electric boiler operating at full load will reduce the overall electric power by C. EB The heating power increases by η EB C EB When the system's peak-shaving capacity is insufficient, the electric boiler is started to consume electricity for heating (F"'F'), reducing the unit's heating demand and thus lowering its electrical output (F"F"'), achieving double the peak-shaving capacity (F"F'). This method only improves the peak-shaving capacity; the compensation heat supply comes from both cogeneration heating and the corresponding cogeneration power generation heat conversion, resulting in the highest peak-shaving cost. The operating constraints of the electric boiler are:

[0140]

[0141] In the formula: The power consumption of the boiler supplied to unit i during time period t.

[0142] All of the aforementioned resources can achieve peak-shaving effects, but their performance (range and direction of regulation, reserve capacity) and economic efficiency (investment cost per unit regulation capacity, peak-shaving cost) differ. Furthermore, variations in system electricity / heat demand, power supply structure, and uncertainties in renewable energy sources lead to different peak-shaving and reserve requirements. The key to collaborative planning lies in how to coordinate and allocate various resources according to system needs to achieve optimal system economics while ensuring safe operation.

[0143] 2. Lightweight and robust collaborative programming model for electrical and thermal flexibility resources considering the risk of insufficient reserves.

[0144] Because wind power output is uncertain, and this uncertainty intensifies when wind power accounts for a high proportion of total power output, flexible allocation of electricity and heat resources increases wind power consumption and reduces energy costs. However, the increased uncertainty of wind power leads to greater safety risks, and different resources have different impacts on these risks. Therefore, the objective function of collaborative planning should consider both economic efficiency and the degree to which system safety risks are reduced. Traditional robustness considers the worst-case scenario with high robustness, but the system faces lower risks and poor economic efficiency. Lightweight robustness can balance economic efficiency and risk loss. Although there are some risks, current systems have mechanisms such as demand-side response to mitigate them, preventing serious safety risks. Therefore, based on the idea of ​​decomposition and coordination, a collaborative lightweight robust planning model for electricity and heat flexibility resources is constructed, which includes investment decisions, day-ahead deterministic scheduling, and quantification of reserve shortage risks that take into account uncertainty.

[0145] 2.1 Electric Heating Flexibility Resource Investment Decision Model

[0146] The investment decision model uses the annual investment cost C of newly added resources at the planning level. INV Annual incremental operating cost C INC-OP Annual reserve shortage risk cost C RISK The goal is to minimize the sum of the three parts.

[0147]

[0148] The decision variables for this model are flexible resource allocation options, including whether or not unit i undergoes low-pressure cylinder cutoff and high back pressure modification. and Thermal storage and electric boiler configuration capacity and

[0149] Therefore, the annual investment cost C of the newly added resources in the objective function INV It consists of four types of flexible transformation methods and resource costs.

[0150]

[0151] In the formula: The investment cost recovery factor for the renovation of Γ is Γ={B,LC,HA,EB}, where α is the discount rate and κ is the return on investment. Γ The service life of the unit is determined by the modification method. EX For the collection of extraction steam turbine units; and G Γ The costs of various modifications to unit i and the unit capacity (MWh / MW) cost of the configured equipment are respectively.

[0152] Annual incremental operating cost C INC-OP This is the difference in annual operating costs before and after the renovation. Because the operating cost before the renovation was C... OP0 This is a constant term that does not affect the optimization results and can be ignored.

[0153]

[0154] Where: N type Typical number of days; and d n These represent the typical daily operating cost and number of days, respectively. The calculation method will be explained in detail later.

[0155] Annual reserve shortage risk cost C RISK The calculation formula is:

[0156]

[0157] In the formula: β is the risk penalty preference. The larger the value, the more the planning model values ​​the potential loss caused by the risk and adopts the risk avoidance planning scheme. In this invention, β is preferably 1. The typical daily reserve shortage risk cost consists of two parts: intraday adjustment cost and risk penalty. The specific calculation will be explained in detail later.

[0158] Typical daily wind power output during the n-hour period t with uncertainties It can be expressed in the form of a box-type indeterminate set:

[0159]

[0160] In the formula: and These represent the maximum and minimum fluctuation ranges of wind power, respectively. and A 0-1 integer variable representing whether wind power fluctuates; Π w The uncertainty limit is preferably 16.

[0161] 2.2 Deterministic Optimization Scheduling Model

[0162] The model aims to determine the typical daily operating cost. This includes the power generation costs (including coal consumption and carbon emissions) of extraction steam and pure condensing units.

[0163]

[0164] The decision variables mainly include the status and output of each unit / equipment at each time period on typical days, as well as wind power output. Specifically, for typical day n, during time period t, the output and operating status of condensing unit i. and The startup status indicators of the extraction steam turbine unit, its heating and power generation output, and the cylinder cut-off status of the unit after the low-pressure cylinder removal modification. and γ LC,i,n,t The extraction back-pressure unit's heating and power generation output and start-up status. and And the thermal storage output, thermal storage capacity and electric boiler output of extraction steam unit i and and wind power output The formulas for each term of the objective function are:

[0165]

[0166]

[0167] In the formula: ρ is the unit coal price; ρ is the equivalent coal price calculated from water, environmental protection, and other costs incurred during unit operation, excluding coal costs, preferably 1.15; τ is the length of a unit time period, preferably 1 hour; γ car Ω represents the unit carbon emission cost; θ is the standard coal CO2 emission coefficient, taken as 0.68; PU For pure condensing units; T is the typical number of hours in a day, which is 24 hours; F Υ,i,n,t and These represent the coal consumption for power generation and the coal consumption for start-up, respectively. Υ={CHP-Λ,CON} is the set of unit types, and Λ={φ,B} is the set of various extraction steam units.

[0168] In addition to the resource operation constraints for typical day n represented by formulas (1) to (10), the constraints of this model also include the following constraints:

[0169] 1) Operating constraints of pure condensing units

[0170] The constraints mainly include the limit output, ramping, shortest start-stop time constraints, and coal consumption function.

[0171]

[0172]

[0173]

[0174]

[0175] In the formula: and These represent the minimum and maximum output of unit i, the upper and lower limit ramp rates, and the shortest start-stop time, respectively.

[0176] 2) Operating constraints of extraction steam turbine units

[0177] To ensure heat load supply during the heating season, extraction steam turbine units should be fully operational during the heating season. During the non-heating season, they can participate in start-up and shutdown scheduling, and must meet the start-up and shutdown time constraints (the same constraints as for pure condensing units). In addition, these constraints also include limit ramp-up constraints and coal consumption functions.

[0178]

[0179]

[0180]

[0181]

[0182] Where: M sea This is a seasonal indicator; it is 1 for the heating season and 0 for the non-heating season. and These represent the uphill and downhill ramp rates of extraction steam turbine unit i under pure condensing conditions; a i and b i These are the marginal coal consumption rate and no-load coal consumption of the extraction steam turbine unit under pure condensing conditions, respectively.

[0183] 3) Heating balance constraints in each heating area

[0184] Heating is balanced by zone based on thermal power plants.

[0185]

[0186] Where: Ω EX,l This refers to the collection of all heating units in a thermal power plant. The heat load borne by a thermal power plant during a typical day n-hour period t.

[0187] 4) System power supply balance constraints

[0188]

[0189] In the formula: and This represents the wind power output and system electrical load during a typical day, n-time period t.

[0190] 5) Wind power output constraints

[0191]

[0192] 2.3 Intraday Rescheduling Model Considering Wind Power Uncertainty

[0193] To mitigate the impact of wind power output uncertainty, based on the day-ahead unit configuration and cylinder cut-off status, intraday reallocation of various units, thermal storage, and electric boilers is conducted to calculate the resource allocation scheme for thermal storage reserve capacity and mitigate the risk of insufficient reserve. This risk penalty is characterized by the cost of wind curtailment and load shedding, which is included as part of the intraday reallocation objective function, with the other part being the cost of additional power generation from the units.

[0194]

[0195]

[0196] In the formula: C wp C loss , and The decision variable x includes the penalty cost per unit power for wind curtailment and load shedding, as well as the amount of wind curtailment and load shedding. In addition to wind curtailment and load shedding, the decision variable x also includes intraday rescheduling variables for various types of generating units, as well as thermal storage and electric boilers.

[0197] In addition to the constraints in formulas (1)-(10), (19), (20), (22), (25), and (26), the model also includes:

[0198] 1) Thermal power balance constraint

[0199]

[0200] 2) Power balance constraints

[0201]

[0202] 3) Wind curtailment constraints

[0203]

[0204] 4) Emergency load shedding constraints

[0205]

[0206] 2.4 Two-Phase Lightly Robust Programming Model

[0207] The two-stage lightweight robust programming model of equations (1)-(35) can be expressed as a two-stage, three-level optimization problem of min-max-min:

[0208]

[0209] In the formula: h(x) represents the constraints of the first-stage investment-day-ahead deterministic scheduling problem, including equations (1)-(10) and (19)-(29), which do not contain uncertain parameters; g(x,u,y) and l(x,u,y) are the inequality and equality constraints of the rescheduling problem under the second-stage uncertain scenario, respectively; Ω(x,u) represents the feasible region of the running variable y given the planning variable x and the uncertainty set u, and its specific expression is:

[0210]

[0211] Equation (37) is a compact form of the optimization problem, where: D, K, F, G, I u A u H u Let be the coefficient matrix corresponding to each constraint, and d and h be the constant column vectors corresponding to the constraint matrices; γ, λ, ν, π, and σ are the dual variables corresponding to each constraint.

[0212] 3. Model Solving Method Based on CCG Algorithm

[0213] For the two-stage lightweight robust planning model mentioned above, this invention uses the CCG algorithm to divide the model into a main problem and sub-problems, and iterates to find the optimal solution. The main problem solves the first-stage variables, namely the planning and the day-ahead deterministic scheduling scheme, while the sub-problems find the worst-case scenario that maximizes the risk of insufficient reserves under the given planning and day-ahead scheduling schemes. By further decomposing equation (37), the objective function and constraints of the main problem are obtained as shown in equation (38):

[0214]

[0215] In the formula: c1 and c2 are the coefficients of the decision variables of the objective function in the first and second stages, respectively; k is the current iteration number of the algorithm, and y l The solution to the subproblem after l iterations is the main problem and the subproblem's solution. Let be the value of the uncertain variable u in the worst-case scenario obtained after l iterations.

[0216] The objective function of the decomposed subproblem is shown in Equation (39). Considering that the max-min bi-level optimization is an NP-hard problem and difficult to solve, but when the first-stage variable x and the uncertain variable u are given, the inner min problem is a linear deterministic problem. According to the strong duality theory, it is transformed into the max form, as shown in Equation (40).

[0217]

[0218]

[0219] Equation (15) can be written as u = u + B + Δu+ -B - Δu - When substituted into equation (40), the product of a continuous variable and a 0-1 variable will appear. An auxiliary variable is introduced to linearize it.

[0220]

[0221] In the formula: B' + B' - C' + C' - M is a sufficiently large positive real number and is introduced as a continuous auxiliary variable.

[0222] After obtaining a robust and feasible day-to-day scheduling scheme during the planning phase, the actual values ​​of the uncertain information are also known. Therefore, the simulation phase becomes a deterministic optimization problem of solving rescheduling variables under uncertain scenarios, which can be expressed as:

[0223]

[0224] The flowchart of the two-stage lightweight robust planning algorithm for electrothermal flexibility resources of this invention is as follows: Figure 4 As shown, the specific steps in the planning phase are as follows:

[0225] Step 1) Generate a set of predicted wind power output values ​​u k As the worst-case scenario variable for the model, we set the lower bound of the total cost corresponding to the final planning scheme as LB = –∞, the upper bound as UB = +∞, and the number of iterations as k = 1;

[0226] Step 2) Determine the worst-case scenario Substituting into the first-stage model equation (38), the corresponding planning and operation results are obtained. k is the iteration number, which is 1 in this case. The calculated comprehensive cost is used as a lower bound of the model and assigned to it.

[0227] Step 3) Substituting the fixed value into the two-stage model equation (41) yields the objective function value under the uncertain subproblem scenario. and the value of the uncertain variable u under the corresponding worst-case scenario. Update Upper Realm

[0228] Step 4) Determine if the difference between UB and LB is less than the given convergence threshold ε. If it is less, output the optimal solution at this time. Otherwise, let the iteration number k = k + 1, jump to step 2) to continue solving, and increase the variable y. k+1 With the following constraints:

[0229]

[0230] This embodiment is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for lightweight and robust collaborative optimization of electrothermal flexibility resources, comprising the following steps: Step 1: Establish a three-layer, two-stage, lightweight robust programming model that includes investment decision-making, day-ahead deterministic optimization scheduling, and intraday rescheduling that takes into account wind power uncertainty. The three-layer, two-stage lightweight robust planning model established in step 1 includes the annual investment cost of newly added resources at the planning level. Annual incremental operating costs Annual reserve shortage risk cost An investment decision-making model is established with the objective of minimizing the sum of the three parts; using a typical day Operating costs A day-ahead deterministic optimization scheduling model is established for the objective; an intraday rescheduling model is established considering the uncertainty of wind power, whose objective function includes risk penalties represented by wind curtailment and load shedding penalties, and the cost of additional power generation by the generating unit; the annual investment cost of the additional resources. It consists of four types of flexible transformation methods and resource costs; the annual incremental operating cost It is the difference in annual operating costs before and after the renovation; the aforementioned annual reserve shortage risk cost The calculation formula is: ； In the formula For risk-punishment preference, For typical days The risk cost of insufficient reserves consists of two parts: intraday adjustment costs and risk penalties. Typical number of days; For typical days Number of days of operation; Typical day in the day-ahead deterministic optimization scheduling model Operating costs The power generation costs, including extraction steam and pure condensing units, encompass coal consumption and carbon emissions; the decision variables for this model include: typical day Time period condensing unit Output and power-on status and extraction steam unit Start-up status indicators, heating and power generation output, and cylinder cut-off status of the unit after low-pressure cylinder cut-off modification. , , and extraction back pressure unit Heating, power generation output and start-up status , and and extraction steam turbine units The power output of the thermal storage, the amount of thermal storage, and the power output of the electric boiler are all provided. , and and wind power output The objective function can then be expressed as: ； ； ； In the formula, Price per unit of coal; The equivalent coal price, calculated by converting water, environmental protection, and other costs incurred during unit operation (excluding coal costs), is 1.

15. The unit time period is 1 hour; Cost per unit of carbon emissions; The standard coal CO2 emission factor is taken as 0.68; A collection of pure condensing units; This represents a typical number of hours within a day, or 24 hours. and These are coal consumption for power generation and coal consumption for start-up. A set of unit types, It is a collection of various extraction steam turbine units; Step 2: Represent the three-layer, two-stage lightweight robust programming model established in Step 1 as a two-stage, three-layer optimization problem of min-max-min; Step 3: Use the CCG algorithm to divide the two-stage, three-layer optimization problem in Step 2 into a main problem and sub-problems, and iterate to find the optimal solution. The main problem solves the first-stage variables, namely planning and day-ahead deterministic scheduling schemes. The sub-problems find the worst-case scenario that maximizes the risk of insufficient reserves under given planning and day-ahead scheduling schemes.

2. The method for lightweight and robust collaborative optimization of electrothermal flexibility resources according to claim 1, characterized in that: The four types of flexible modification methods include low-pressure cylinder removal modification, high back pressure heating modification, configuration of thermal storage and configuration of electric boilers; the difference between the compensation heating cost of each resource and the original unit's direct co-production heating cost is used as the measurement standard for peak-shaving cost, and coal consumption is its unit of measurement. The low-pressure cylinder cut-off modification refers to cutting off the steam inlet of the low-pressure cylinder of the extraction steam unit, keeping only a small amount of cooling steam, so that the low-pressure cylinder operates under high vacuum conditions with "zero output". The power generation reduction and the heating power increase after the cut-off are constant values, with double the peak-shaving space. When the cylinder is cut off, the steam used to do work in the low-pressure cylinder is completely extracted for heating, which is still cogeneration heating, and the peak-shaving cost is 0. The high back pressure heating retrofit refers to the turbine operating at a higher back pressure during the heating season, using the exhaust steam from the high, medium, and low pressure cylinders to generate electricity and then heating the heating network water. The heating network water then enters the heating network heater and is supplemented by extracted steam for heating, thus achieving decoupling of the heat and electricity components. The heating peak-shaving capacity is enhanced, and the power supply regulation range is reduced and narrowed. The exhaust steam increases the unit's heating capacity. According to the "benefit goes to electricity" method, the high back pressure heating peak-shaving cost is negative. The described thermal storage configuration involves connecting the thermal storage device in parallel with the heating unit. When there is surplus heat from the combined heat and power generation of the steam turbine, the thermal storage is stored. When the power output of the steam turbine needs to be reduced or increased for peak shaving, the thermal storage is released to supplement the insufficient heat supply from the steam turbine. This configuration also enhances peak shaving capabilities and provides backup capacity to cope with the uncertainty of wind power in extreme situations. The heat supply for compensation comes from the combined heat and power generation during other periods, and the peak shaving cost is 0. The electric boiler configuration involves connecting it in parallel with the original heating system of the unit. When the electric boiler operates at full load, the overall power consumption decreases while the heating power increases. When the system's peak shaving is insufficient, the electric boiler is started to consume electricity for heating. The unit's heating demand decreases, causing its power output to decrease, thus achieving double peak shaving and maximizing peak shaving costs.

3. The method for lightweight and robust collaborative optimization of electrothermal flexibility resources according to claim 2, characterized in that: The constraints after the low-pressure cylinder removal and modification include: 1) Feasible operating range constraints: ； ； ； In the formula, 0-1 variables and These represent extraction steam turbine units. Intraday session Engine start-up and cylinder cut-off status , and , These represent the combined heat and power (CHP) power supply and heat power components of the low-pressure cylinder cut-off retrofitted unit in the non-cut-off and cut-off cylinder sections, respectively. , Indicates the overall electrothermal output of the unit; 2) Daily low-pressure cylinder cutoff frequency constraint ； In the formula, The scheduling cycle is within the day; The constraints after the high back pressure heating system renovation are: 1) Modify the start-up operating range constraints: ； In the formula, 0-1 variables and They are extraction steam turbine units Whether or not the high back pressure heating system has been upgraded and its operating status. and These are extraction back pressure units The electrothermal ratio when operating under pure back pressure and maximum extraction steam conditions; and The units Maximum and minimum power generation under pure back pressure conditions; For the unit Maximum heating capacity; 2) Unit ramp rate constraint: The ramp rate constraint of back-pressure units is represented by the power ramp rate constraint under pure back-pressure conditions with a slope of -1 on the equal coal consumption line. ； In the formula, and The units Uphill and downhill gradient rates; for Time-of-use units Electric and heat power are converted to the electric power under pure back pressure conditions under the equal coal consumption line; 3) Unit coal consumption function: The coal consumption of a back-pressure unit is expressed using the coal consumption under pure back-pressure operating conditions on an equal coal consumption line. ； The thermal storage charging and discharging power and energy constraints of the configured thermal storage method are as follows: ； ； In the formula: , and unit thermal storage Time-limited heat storage capacity and maximum charging and discharging capacity; For heat storage efficiency; For capacity, and These are the maximum heat storage and release power of the thermal storage device; The operating constraints of the electric boiler are: ； In the formula: The electric boiler supplied to unit i Power consumption during a given time period; , This refers to the capacity and electro-thermal conversion efficiency of the electric boiler.

4. The method for lightweight and robust collaborative optimization of electrothermal flexibility resources according to claim 1, characterized in that: The aforementioned deterministic optimization scheduling model includes operating constraints for pure condensing units, operating constraints for extraction steam units, operating constraints after the configuration of four types of flexibility modification measures, heating balance constraints for each heating area, system power supply balance constraints, and wind power output constraints.

5. The method for lightweight and robust collaborative optimization of electrothermal flexibility resources according to claim 4, characterized in that: The operating constraints of the pure condensing unit include limit output, ramp-up, shortest start-up and shutdown time constraints, and coal consumption function. ； ； ； ； In the formula, , , , , and The units Minimum and maximum output, upper and lower limit climbing speed and shortest start-stop time; The extraction steam turbine units should be fully operational during the heating season, and may participate in start-up and shutdown scheduling during the non-heating season, provided that start-up and shutdown time constraints are met. Additionally, the operational constraints for the extraction steam turbine units include limit ramp-up constraints and a coal consumption function. ； ； ； ； In the formula, This is a seasonal indicator; it is 1 for the heating season and 0 for the non-heating season. and They are extraction steam turbine units The uphill and downhill gradient rates under pure condensation conditions; and They are extraction steam turbine units Marginal coal consumption rate and no-load coal consumption under pure condensing conditions.

6. The method for lightweight and robust collaborative optimization of electrothermal flexibility resources according to claim 4, characterized in that: The heat supply balance constraints of each heating zone are as follows: ； In the formula, For thermal power plants All heating units are combined; For thermal power plants typical day Time period The heat load it bears.

7. The method for lightweight and robust collaborative optimization of electrothermal flexibility resources according to claim 4, characterized in that: The system power supply balance constraint is: ； In the formula: and For typical days Time period Wind power output and system electrical load.

8. The method for lightweight and robust collaborative optimization of electrothermal flexibility resources according to claim 5, characterized in that: The objective function of the intraday rescheduling model, which takes into account the uncertainties of wind power, is: ； ； In the formula, , , and This includes the penalty cost per unit power for wind curtailment and load shedding, as well as the amount of wind curtailment and load shedding; excluding wind curtailment and load loss. and In addition, decision variables This also includes intraday rescheduling variables for various generating units, as well as thermal storage and electric boilers. , , , , , , , .

9. The method for lightweight and robust collaborative optimization of electrothermal flexibility resources according to claim 8, characterized in that: The intraday rescheduling model includes four types of operational constraints after the configuration of flexibility modification measures: thermal power balance constraints, electrical power balance constraints, wind curtailment constraints, and emergency load shedding constraints.

10. The method for lightweight and robust collaborative optimization of electrothermal flexibility resources according to claim 9, characterized in that: The thermal power balance constraint is ； The power balance constraint is ； The wind curtailment constraint is ； The emergency load shedding constraint is as follows: 。 11. The method for lightweight and robust collaborative optimization of electrothermal flexibility resources according to claim 1, characterized in that: The two-stage, three-layer optimization problem is: ； In the formula: The constraints are for the first phase investment-day-ahead deterministic scheduling problem, and the formula does not contain uncertain parameters; , These are the inequality and equality constraints for the rescheduling problem under the uncertain scenario in the second stage; Represents a given planning variable With the uncertain set At that time, run variables The feasible area.

12. The method for lightweight and robust collaborative optimization of electrothermal flexibility resources according to claim 11, characterized in that: In step 3, the two-stage, three-layer optimization problem is decomposed to obtain the objective function and constraints of the main problem: ； In the formula: , These are the coefficients of the decision variables for the objective functions in the first and second stages, respectively. This represents the current iteration number of the algorithm. The main and sub-problems are in the process of Solution to the subproblem after the next iteration For the process Uncertain variables in the worst-case scenario obtained after the second iteration The value of ; The objective function of the decomposed subproblems is: ； Based on strong duality theory, the objective function of the subproblem is transformed into a max form, which is: ； Typical day Time period Uncertain wind power output Substituting the transformed subproblem objective function into the linearized function using auxiliary variables, we obtain... ； In the formula: , , , For the introduction of continuous auxiliary variables, It is a sufficiently large positive real number.

13. The method for lightweight and robust collaborative optimization of electrothermal flexibility resources according to claim 12, characterized in that: The specific process of the iterative loop in step 3 is as follows: Step 1) Calculate a set of wind power output forecast values As the worst-case scenario variable for the model, we set the lower bound of the total cost corresponding to the final planning scheme as LB=–∞, the upper bound as UB=+∞, and the number of iterations as k=1; Step 2) Determine the worst-case scenario Substituting these values ​​into the objective function of the main problem, we obtain the corresponding planning and execution results. , Let be the number of iterations, at this point The value is set to 1, and the calculated comprehensive cost is used as a lower bound of the model and assigned to LB= ; Step 3) Substituting the fixed values ​​into the linear expression of the objective function of the decomposed subproblem, we obtain the objective function value under the uncertain scenario of the subproblem. and the corresponding uncertainties in the worst-case scenario The value of Update the upper bound UB= ; Step 4) Determine whether the difference between UB and LB is less than the given convergence threshold. If it is less than, then output the optimal solution at this time. Otherwise, reduce the number of iterations. Proceed to step 2) to continue solving, and add variables. With the following constraints: 。

Images